From 1989 through 1991, 54 consecutive tubal pregnancies were managed by laparotomy by two surgeons. Serum and peritoneal fluid samples were collected to measure levels of human chorionic gonadotropin (hCG). One surgeon injected vasopressin into the mesosalpinx and the bed of the implantation site for hemostasis and the other did not. The two patient populations were similar with respect to gestation age and size, and serum and operative peritoneal hCG levels. Of the 42 women in whom hCG ratios were available, only 2 of 20 with initial serum titers less than 2000 had ratios less than 2. The peritoneal:serum hCG ratio is valuable in making a diagnosis of ectopic pregnancy when ultrasound is unreliable. There was no difference in the regression slopes of postoperative hCG or in hematocrit changes between the two groups. Of the 54 women, only 1 developed a persistent ectopic pregnancy; she did not receive vasopressin. Vasopressin is effective in managing hemostasis. The relative local hypoxia does not seem to affect the viability of remaining trophoblasts as measured by the rate of disappearance of hCG.